Date and Time: Tuesday, June 16, 2026, 8:00–11:00 AM
Location: UCI Anteater Learning Pavilion (ALP), Irvine, California
Format: Half-day workshop
See ISLS 2026 Workshop Listing.
As AI rapidly enters education, learning scientists are being called upon to guide what constitutes educative AI rather than merely productive AI. Yet influence requires more than critique—learning scientists can make a difference through participation in the design of AI infrastructure itself: open datasets, models, benchmarks, and data governance mechanisms that embed human learning principles from the ground up. The overarching theme is to explore how the learning sciences can be systematically integrated into infrastructural “public goods” for AI in education, seeking both to understand how learning scientists could more strongly get involved in this work while also retaining a critical stance and naming concerns that must be addressed.
This workshop seeks to collect insights on the why, what and how of integrating learning sciences into AI resources. It will facilitate knowledge networking, sharing and discussion, while encouraging critical examination of safety, privacy, ethics, and equity concerns.
Speakers
Jeremy Roschelle
Rene Kizilcec
Tyler Sussman
Short paper presentations: approximately 8 design opportunity cases submitted by the community.
Collaborative design groups: cross-sector teams co-develop infrastructure concept cards specifying (a) the learning construct embedded, (b) the infrastructural target (dataset, model, benchmark), and (c) expected ethical or equity implications.
Synthesis plenary: collective brainstorming of key themes for an ISLS Brief (approximately 2,000 words) on Learning-Sciences-Informed AI Infrastructure for publication on the ISLS website. Participants identify next steps, collaborations, and shared resources
Town Hall: How could we involve more people in the Learning Sciences in this work, while also retaining our critical stance?
Participation Requirements and Selection
Participation will be open. Applicants will complete a brief survey (available soon) describing:
- Their professional background and relevant expertise
- Brief statement of interest in infrastructure for AI
- Current or planned work related to learning sciences and AI
- Geographic location and institutional affiliation
Selection will prioritize diversity in geography, discipline, career stage, and representation from research, practice, and industry. Priority given to applicants who can commit to both virtual pre-sessions and in-person workshop. A minimum of 8 participants is required; optimal participation is 25–40.